"We Use AI" Is No Longer a Differentiator: Zerodha Founder's Warning to Startups

Zerodha's Nithin Kamath warns AI alone is no longer a differentiator — startups must demonstrate genuine competitive moats beyond simply integrating AI into their products.

"We Use AI" Is No Longer a Differentiator — that is the blunt message from Nithin Kamath, co-founder and CEO of Zerodha, India's largest retail stockbroking platform, as he challenges the next generation of Indian startup founders to articulate genuinely defensible competitive advantages that go far beyond slapping an AI label on their pitch decks.

Kamath's remarks, shared during a high-profile fireside chat at the TiE Bengaluru Summit 2026, strike at a fundamental shift in how investors and market incumbents evaluate early-stage startups. In 2023 and 2024, merely integrating a GPT wrapper or LLM-powered chatbot was enough to attract seed-stage interest. By mid-2026, AI has become infrastructure — as commoditized as cloud hosting or mobile-first design.

The AI Commoditization Inflection Point

Every single pitch deck I've seen in the last six months says 'AI-powered.' That's like saying 'we use electricity.' It tells me nothing about your defensibility,

Kamath said to an audience of over 1,200 founders. "The question I now ask is: what happens when your competitor — who also uses the same OpenAI or Gemini API — launches in three months? What is your moat?"

The observation reflects a measurable trend in Indian venture capital. According to Tracxn's Q2 2026 India Startup Report, the number of Indian startups self-describing as 'AI-first' grew 340% between 2023 and 2026, but median seed-stage valuations for generic AI wrapper companies fell 28% over the same period. Investors are explicitly discounting AI as a feature and pricing it as baseline expectation.

#

Funding Trends: AI-Native vs. AI-Enabled Startups

| Metric | AI-Native Startups (Proprietary Models) | AI-Enabled Startups (API Wrappers) | Non-AI Startups | | :--- | :--- | :--- | :--- | | Median Seed Valuation (2026) | $12M – $18M | $4M – $7M (↓28% YoY) | $5M – $9M | | Series A Conversion Rate | 34% | 11% | 22% | | Investor Interest Score | 8.7 / 10 | 4.2 / 10 | 6.1 / 10 | | 12-Month Survival Rate | 78% | 41% | 58% |

What Actually Constitutes a Moat in 2026?

Kamath outlined five categories of genuine defensibility that he looks for when evaluating startups as an angel investor:

#

1. Proprietary Data Flywheels

Startups that generate unique, self-reinforcing datasets through user interactions build compounding advantages that no API call can replicate. Zerodha itself exemplifies this: its Kite platform processes over 15 million daily orders, generating proprietary market microstructure data that feeds increasingly sophisticated risk management and order routing algorithms.

#

2. Vertical-Specific Domain Depth

Horizontal AI tools face existential competition from OpenAI, Google and Anthropic themselves. Vertical specialists — companies building deeply domain-specific models for Indian agriculture, vernacular content or regulatory compliance — create natural moats through specialized training data and regulatory expertise.

The winners will be founders who pick a narrow vertical and go impossibly deep. Build the model that understands Indian GST compliance better than anyone — not another generic summarizer,

Kamath emphasized.

#

3. Distribution and Network Effects

Product distribution channels — partnerships with banks, integration into government platforms like ONDC, embedded workflows in enterprise procurement stacks — create switching costs that pure technology advantages cannot.

#

4. Unit Economics Discipline

Kamath, who famously built Zerodha to profitability without external venture capital, stressed that sustainable unit economics remain the ultimate moat: "If you're burning ₹3 to earn ₹1, no amount of AI makes that a business. AI should reduce your cost-to-serve, not increase your marketing burn."

#

5. Regulatory and Compliance Moats

In heavily regulated sectors — fintech, healthtech, edtech — obtaining licenses, certifications and compliance clearances creates durable barriers that pure-play AI startups often underestimate.

The Broader Market Context

Kamath's warning arrives as Indian startups continue to raise significant capital — $13.4 billion across 1,240+ rounds in 2026 so far. However, the composition of funded companies is shifting decisively. Investors are channeling capital toward startups with defensible proprietary technology, proven distribution and clear paths to profitability rather than toward generic AI-wrapper plays.

The shift also aligns with the broader industry conversation around India moving from AI adoption to AI ownership, where the emphasis is on building foundational technology rather than simply deploying others' tools.

Key Takeaway for Founders

The message is clear: AI is now table stakes. In 2026 and beyond, the startup founders who win will be those who can answer one simple question — if we strip away the AI label, what remains that is uniquely defensible? Those who cannot answer convincingly will find themselves competing in an increasingly commoditized market where margins compress and differentiation evaporates.